The world is seeing great progress in development of new treatments for diseases. However, there are many diseases where effects of interventions are difficult to measure and where the clinical meaningfulness is hard to evaluate based on observed treatment differences in any set of standard measures. 

In complex progressive diseases, such as Alzheimer’s disease, the next generation of treatments are unlikely to rescue lost function or fully stop the disease progression but may slow the progression of disease. However, the decade-long course of Alzheimer’s disease combined with the impracticality of trials of many years’ duration may lead to treatments showing only small benefits on typical cognitive outcome measures in trials. The clinical meaningfulness of such treatment effects can be questioned, but even small benefits may represent substantial slowing of disease progression that could translate into long-term benefits. 

In a new article published in Statistics in Medicine, Lars Lau Raket proposes a novel class of models to estimate and quantify treatment effects in the situations outlined above. These so-called Progression Models for Repeated Measures (PMRMs) can quantify treatment effects in terms of slowing or time delays observed on typical continuous outcome measures used in trials in progressive diseases. Compared to conventional estimates of treatment effects where the unit matches that of the outcome scale (e.g. 2 points benefit on a cognitive scale), these time-based treatment effects can offer better interpretability and clinical meaningfulness (e.g. 6 months delay in progression of cognitive decline). Furthermore, the common unit of time on these treatment effects enable better comparison of treatment effects across difference outcome measures and across trials. 

In the article, different PMRMs are explored and compared to conventionally used approaches. In addition to the more straightforward interpretability of these treatment effects compared to conventional methods, the PMRM models are shown to increase the power to detect disease-modifying treatment effects where accumulating benefits are seen over time.